Himalayan Journal vol.07
The Himalayan Journal

Publication year:

Kenneth Mason
    (H.W. Tilman)
    (I. GENERAL r. finsterwalder)
  5. A VISIT TO NUN KUN, 1934
    (F. S. SMYTHE)
    (J. B. AUDEN)
    (A.P.F. Hamilton)
    (Captain F. KINGDON WARD)
    (P. C. DUNCAN)
    (Lieut.- Colonel Kenneth Mason)
  16. NOTES


Lieut.- Colonel Kenneth Mason

The publication of Burrard's and Hayden's standard summary in 1907-8 came at a time when there was a danger that geographers might lose their way in a maze of unclassified detail. It marked a most important stage in the geographical study of the Himalaya. As the result of a hundred years of activity by the Survey of India, we had maps of the greater part of the Himalaya, though the survey of the more inaccessible parts and the regions beyond had been admittedly generalized for reasons of cost, and the maps of Nepal were compiled from the rough route surveys and reports of native explorers, for political reasons. The work was mainly a centenary review, rendered necessary, as stated in the Preface, by the realization that 'the number of travellers in the Himalaya and Tibet is increasing and a wider interest is being evinced by the public in the geography of these regions'. The book was a collection of the known facts, chiefly from reports of the Survey of India, and certain proposals for a systematic and simple classification of the ranges, based mainly on the results of the trigonometrical Survey, were made.

This first volume was a masterpiece of lucidity, and there is little doubt that many of the advances made in Himalayan geography must be traced directly or indirectly to the stimulus it provided. Since 1907 there has been an enormous increase in our geographical knowledge. One of the earliest acts of Sir Sidney Burrard as Surveyor General was to initiate a complete new topographical survey of the Himalaya west of Nepal, on the one-inch scale, carefully contoured and showing clearly the areas of forest and cultivation; and though, owing to the Great War and other causes, mostly concerned with finance, the whole survey is far from complete, we have a good modern one-inch survey of the Himalaya in Kumaun, Kulu, and Kashmir; there exists a good reconnaissance survey of much of the eastern Himalaya, and more recently, by the energetic initiative of Sir Edward Tandy, and under the general direction of his brother, Colonel M. O'C. Tandy, the whole of Nepal has been officially surveyed on the quarter-inch scale, though not by British officers or under their supervision in the field. West of the Indus a complete modern survey has been made of almost the whole of Dir, Swat, and Ghitral, and of the neighbouring districts of the Gilgit Agency, much of which was previously only roughly mapped. Since 1907 also, numerous private expeditions, led by experienced mountaineers and scientists from abroad, and often including geologists, surveyors, and other officers from scientific departments in India, have travsered the Himalaya, the Karakoram, and parts of Tibet, have made detailed maps of the more inaccessible regions, and have, with the aid of modern maps, made scientific observations of the first importance. The time had undoubtedly arrived when another authoritative summary was required.

The task of revising the geographical sections was taken up by Sir Sidney Burrard, the original author; the late Sir Henry Hayden's geological chapters have been brought up to date by Dr. A. M. Heron, of the Geological Survey, who acknowledges the help he has received from other members of his department; and the revised edition became available in separate Parts during the years 1933 and 1934.

The revision of what had become the standard book on the subject was a very formidable undertaking, much more formidable than the preparation of the original work. Owing to the greatly increased public interest in the Himalaya during recent years, to the greater and more varied calls on scientific officers' time in India, and to retrenchment of posts and money, much of the material available for revision is no longer in the records of a single Government department, but is to be found in British and foreign scientific journals and other publications. In 1907 Sir Sidney Burrard was able to consult the records which were in his care, and he was in fairly close touch with explorers; for some years since his retirement in 1919 persistent ill-health prevented that close touch being maintained; and in 1930, when he began the revision, he was separated both by time and distance from the records and the travellers he had to deal with. Access to Indian maps in England is expensive and not always easy, owing to the secretive habit of publishing frontier maps 'for official use only',1 and under modern conditions in India it is not easy for officers to find time for the necessary research. It would undoubtedly have been sounder policy to place an officer in India on special duty to assist the author. The Survey of India claims the right to be the 'geographical authority' on India and the Himalaya, now as in the past. Such a claim carries responsibilities. If it is to be maintained in the future, the Government of India must supply the funds necessary for the department to carry out those responsibilities, and the Surveyor General must encourage his officers to be something more than merely topographical and geodetic surveyors.

1 The maps of Nepal, Northern Baltistan, and Hunza Nagar were all published in this form, though most of them have now been released to the public.

The authors of the revision have handicapped themselves by following the pattern of the first edition, and by writing the new one in the same four parts, in the same order: I. The High Peaks of Asia; II. The Principal Mountain Ranges of Asia; III. The Rivers of the Himalaya and Tibet; and IV. The Geology of the Himalaya. Such an arrangement was suitable to the earlier state of our knowledge; but the time has come when Himalayan geographers must understand the general structure of the mountains before attempting to describe their geography. It may be true, as Sir Sidney Burrard states, that 'in the earlier stages of geographical investigation, the most important features of a mountain mass are the high peaks'. But those stages are now past and with good modern maps and good modern geological surveys much more can be learned from a detailed study of the structure of the ranges, the effects of climate, and the regime of the rivers, than from poring over lists of latitudes, longitudes, and heights of the highest summits. For this reason a complete rearrangement would have been preferable, with the structural sections written afresh as Part I by the geological author, so that the geography could have been built soundly on such a foundation. As it is, the geographical chapters are largely compilations of extracts from the writings of early surveyors and almost ignore the lessons to be derived from the study of structure, or refer to them as too technical and abstruse for the comprehension of the ordinary geographer (p. 244). It may seem ungenerous to criticize a work that has taken so much time and labour. In view, however, of the high place the first edition deservedly held in the literature of the Himalaya, a detailed examination of the new publication appears necessary, especially as it is published under the orders of the Government of India.

Of the seven chapters of Part I in the 1907 edition, the old Chapters 2, 5, and 6 (Notes on certain of the great peaks, the frequency with which certain heights tend to occur, and the geographical distribution of the great peaks) have been omitted, while a single chapter on nomenclature-8 pages of the early edition-has been extended to cover three chapters of 46 pages, often highly controversial, one-sided, and unconvincing. Chapter I contains revised tables of peaks, arranged as before in their orders of magnitude (I-V), with their positions and heights. The advantage of having officially accepted heights and names must be obvious to all, and they should be as complete and authoritative as possible. They should also meet the needs of those who use them, and they should be checked and checked again until there is no possible error in the printed results. They might then be conveniently relegated to an appendix. The new lists include 86 in the first five tables; my own lists which I have kept up to date during the last 25 years now show 117-some no doubt only secondary summits, and hardly true peaks -but still only those peaks of interest to the Survey of India and not all the high peaks of Asia. A number of extremely awkward designations for some of the higher peaks have happily been discarded, and geographers will be grateful. Of the third magnitude peaks (26,000- 27,000 feet), K3, K3a, K, KS,T45, XXX, XXXIX, XXXIV, have all been replaced by group names, such as Gasherbrum III, Anna- purna I. A retrograde step, unexplained till we reach Part III, which deals primarily with rivers, is the substitution of the symbols E1, E2, E3, for the well-known names, Lhotse, Nuptse, and Changtse, which are already well established in the literature of Mount Everest, which appear on the Survey of India maps, which had the approval of the Surveyor General and the Director of Map Publication of the day, and which are far more sensible, far more convenient, and far more acceptable to geographers and mountaineers, for whose co-operation Sir Sidney appeals in a special note. Most of the other naked capitals have gone from the lists, Ts, Bs, Ks, XXXs, XLs, and most of the peaks represented by them can now appear decently clothed in Himalayan literature. One, surely owing to the iniquity of a printer, has put on fancy dress: the peak XLVI (25,064 feet) is disguised as a 'satellite of Gosainthan' though some 170 miles from that mountain. Perhaps B504 (25,134 feet), only two miles from Gosainthan, was measured for this garb, though 'Gosainthan II' would be more appropriate apparel. XLIII (25,429 feet), XLIV (25,271 feet) and XLVI, the offending satellite, are all on the Dhaulagiri Himal, and could be conveniently grouped together, as Dhaulagiri II, III, and IV.1 Then all the peaks of the 4 th magnitude would be decently clothed.

There are a few other points to note about these lists. The positions and heights of the Tirich Mir peaks have all been revised from recent surveys. Some summits omitted from the various lists might have been included: such, for instance, as have been fixed accurately by triangulation for position, and whose heights have been fixed by modern topographical survey within a few feet: Pk 5/42 P (25,250 feet), of the Dasto Ghil group, and Pk 8/42 P (24,860 feet), of the Momhil group, fall within this remark. A few previously accepted peaks have been rightly omitted, having been found by detailed topographical survey to be non-existent. It seems that Kun Lun No. 1 (24,306 feet) should follow their lead, or at least sink to Table VI, where a few doubtful peaks have been placed, for the surveys of Sir Aurel Stein found no peak of this altitude in this region. It is a little difficult to see why Pk 2/78 A (24,244 feet) has been placed in this list of outcasts (Table VI), and why it has been labelled a 'satellite of Kangchenjunga'. According to the data given it has been fixed from five stations, is within one mile of Jonsong peak, which, though only fixed from two stations, is rightly placed in Table V; it is no less than twelve miles from Kangchenjunga, and has been definitely proved to exist. Lastly, Kungur I is out of its correct order in Table IV; with its height of 25,146 feet, it should be four places higher on the list.

1 As suggested in Himalayan Journal, vol. vi, 1934, p. 85.

The new Chapter 2 deals in general with the evolution of geographical names. Chapter 3 is highly controversial and would, in the present writer's opinion, have been better omitted. Since, however, certain opinions are expressed which appear to be very definitely unsound, and must, if adopted, lead to considerable confusion, attention must be called to them. The author explains that when he writes of Tibet he means the country inhabited by all those people who speak the Tibetan language or are of Tibetan stock (pp. 14-16). He states quite truly that the name 'Little Tibet' was a name for Baltistan in the past, quotes early geographers, the Arab Istakhri, Ibn Khurdaba, Marco Polo, Moorcroft, and historians, such as Francke, to support him; he quotes Sir George Grierson to show that the Zoji La is on the 'ethnographic watershed between the Aryan and Tibetan populations' (p. 11); and he states that the ethnographic boundary 'follows an irregular line through Spiti, Garhwal, Nepal, and Sikkim' (p. 11). Presumably therefore all the country north of this 'boundary' should, in the author's opinion, be considered 'Tibet'. Such a conception must lead to endless confusion, as it undoubtedly did at the Royal Geographical Society meeting on the 12th May 1930,1 and subsequently (p. 15). The historical expression 'Little Tibet' for Baltistan, the structural designation 'the Western Tibetan plateau' which extends into Ladakh, the indication on racial and linguistic maps that people of Tibetan stock or speaking the Tibetan language extend beyond the boundaries of political Tibet, are all correct in their respective spheres. But, by itself, the word Tibet means political Tibet to the vast majority of people, geographers as well as others, and the statement that a certain officer worked in Tibet, when in fact he went no farther than Baltistan and Ladakh, is incorrect and misleading. Contrary to the definite implication of the author (p. 15), the name 'Little Tibet' is not shown on modern maps of the Survey of India. It is not shown on the map of 'The Himalaya and Surrounding Regions', published under his direction as Surveyor General, nor on any Survey of India map published during or since his Surveyor-Generalship; nor is it shown on The Times Atlas, the Oxford Atlases, nor on any other modern maps or atlases, other than historical maps or others long since out of date, that I have consulted. On a later page (p. 30), the author, to add force to another argument, writes: 'Maps are scientific documents, used in international agreements.' Would a map, we may ask, with the word Tibet attached to a large part of Kashmir, Spiti, Nepal, and Sikkim be a useful document in international agreement? It would be as wrong as the use of the word 'Arabia' for large parts of Africa, or the use of the colour red for the United States on a map of the British Empire.

1 Geographical Journal, vol. lxxvi, 1930, discussion, p. 157.

Other controversies dealt with in Chapter 3 are those on the various names that have been proposed for Mount Everest, on the origin of the name 'Kangchenjunga', and on Karakoram nomenclature. Mount Everest is so well established in Himalayan literature that there is no fear of it now being displaced; and it seems ungenerous at this stage to decry the brilliant work of Brian Hodgson and to write him down as an unsound geographer (p. 22).1 Hodgson was a great linguist and we have already been told that 'geography and language and history are all parts of one whole' (p. 7); but this was in another connexion. Hodgson's 'unsoundness' apparently lay in the fact that he was convinced that Mount Everest had a Sanscrit name, 'Devadhunga', and we are reminded that the fallacious character of his argument was 'apparent to any person competent to understand the subject'. In fairness to Hodgson, it should be added that as recently as 1932 Dr. Suniti Kumar Chatterji, an eminent and authoritative Bengali philologist, wrote:

If the educated Indian really felt anything about these geographical orthographies, he felt more at Gaurishankar and Deodunga being dethroned by Everest, rather than Kanchana-jangha being proved to be, not Sanscrit and Aryan (and so associated with the gods and rishis and heroes of ancient India) but a parvenu in borrowed Sanscrit garb from the neighbouring land of Tibet {Himalayan Journal, vol. iv, 1932, p. 208).

As regards the name Kangchenjunga, the Tibetan origin of this name has been fully established and accepted by competent Sanscrit and Tibetan scholars. This conclusion was reached after Part I had gone to press, and though the matter is again raised in Chapter 22 dealing with the rivers of the Assam Himalaya (pp. 209-12), the author was too deeply committed to his own view-point to alter it.

Sir Sidney Burrard is no more convincing in his arguments on what he calls the Karakorum controversy',1 mainly through an initial misconception. No one has in the last thirty years suggested abolishing the name 'Karakoram', but, owing to the various applications of the term by different geographers, an effort has been made to secure a suitable nomenclature that will meet modern scientific needs and will meet with general approval. The author does not mention the fact that the Surveyor General, Sir Edward Tandy, in 1928, fully realizing the importance of general agreement on a question of primary nomenclature, officially suggested that the Royal Geographical Society should convene a meeting of competent geographers to consider proposals for the classification of the admittedly complicated mountains of the Karakoram system. Sir Sidney Burrard's memory is at fault when he writes that the proposals were discussed at a meeting in 1926 and that 'Colonel Wood made a protest against them, and he voiced the opinion of Survey officers, political officers, and sportsmen'. What happened was this: Subsequent to a meeting in 1926, in which Colonel Wood took part, certain suggestions were made and published in the Geographical Journal. A year later these suggestions were discussed by officers of the Survey of India conversant with the difficulties of the problem, and obtained the general approval of the Surveyor General, with the result that he issued the invitation to the Royal Geographical Society mentioned above. At the discussion, at which the proposals were put forward, in 1930, not 1926, with the Surveyor General's approval, Colonel Wood was not present, made no protest, and the proposals received the general support of Sir Francis Younghusband, General the Hon. C. G. Bruce, Dr. T. G. Longstaff, and others fully qualified to speak for political officers and sportsmen (Geographical Journal, vol. lxxvi, pp. 143 et seq.).

1 Hodgson has been proved correct in more than one geographical controversy; for instance, his names for the tributaries of the Gandak, though not accepted at the time, are now identifiable on modern maps.

These illustrations of the controversial nature and inaccurate presentation of some of the matter in the book must suffice; but they undoubtedly detract enormously from its value as a serious contribution to the subject.

1 The author spells the word 'Karakoram' throughout the first edition, and 'Karakoram' throughout the second, but gives no reason for the change. Survey of India maps and modern official literature have Karakoram, and Sir Aurel Stein tells me that there is not the slightest doubt that this is the more correct form. There is no question whatever about the pronunciation: the termination is pronounced rum not room, in English, which becomes ram not rum in the Hunterian system of transliteration adopted in India, pronounced like the stuff we drink, and not like the place we drink it in. Hence, if the Hunterian system is applicable for Turki words, Karakoram is correct, and Karakorum is wrong: if we employ 'taphouse’ methods, Karakorum is right, and Karakoram is incorrect. For many years the Survey of India has used Karakoram, and to my certain knowledge the Surveyor General decided less than four years ago to retain that spelling. The decision was communicated officially to Sir F. De Filippi, with the result that he adopted Karakoram in the English edition of his book. The decision was taken after consultation with philological experts in England, and a copy of the correspondence was sent by the Director of Map Publication, Calcutta, to the Editor of the Himalayan Journal to note. We do not therefore intend to change.

Part II of the 1907 edition was a detailed attempt, and in general a successful attempt, to sort out the various peaks triangulated by the Survey of India into ranges, and it became the standard of reference for Himalayan geographers. The principle adopted, in our then limited knowledge of structural detail, was to plot the positions of the high peaks and to assume range alinements where these positions were considered sufficiently close to justify such assumptions. The general alinement of the Great Himalaya and the ranges to the south were, except in Nepal, comparatively well known, and from the results of these plottings, the Great Himalaya between Bhutan and the Indus, and certain lesser ranges and ridges of the Himalaya, were established. In general, these have stood the test of more detailed examination on the ground, though some of the so-called lesser Himalayan 'ranges' are in reality ridges due to differential erosion, and some are due to 'nappe' structure. It is a pity that no attempt has been made to utilize the valuable contributions to these structural problems made by Messrs. Pilgrim, Wadia, Auden, and West of the Geological Survey of India. Moreover, the clue to the structure and orogeny of the north-west Himalaya may be found in Dr. Wadia's important paper on the North-west Syntaxis of the Himalaya,1 to which only the briefest allusion is made on pages 94 and 243, though it affects the structure and geography of all the mountains at least as far as the foot of the Pamirs. The detailed modern maps of the Survey of India confirm Wadia's solution in a most remarkable manner, and one can only wonder that surveyors had not already reached the same conclusion. The stage has long since been reached when Himalayan geographers can afford to wait for geologists to 'be eventually pressed to explain their solution in simple language' (p. 244). Wadia's solution is extremely clear, concise, and to the point. His results are of the greatest importance, yet no notice has been taken of them in the geographical charts.

Nor in the new edition does a sufficient study appear to have been made of the modern maps of Nepal. On some the rivers have been shown incorrectly as in the old edition, though we now know their courses with accuracy.

1 'The Syntaxis of the North-west Himalaya: its Rocks, Tectonics, and Orogeny', by D. N. Wadia, Records, Geological Survey of India, vol. lxv, Pt. 2, 1931, pp. 189-220.

In Chapter 11, which is entitled 'The ranges that separate the Himalaya from Tibet', the author is obviously hampered by his own definition of Tibet, to which reference has already been made, for according to him Tibet reaches the Zoji La, on the Great Himalaya on the west, and elsewhere its southern boundary follows an irregular line south of the great range through Spiti, Nepal, and Sikkim. But, with the curious apology 'and as thus we have to make a topographical compromise that will be in harmony with the ethnological and linguistic conditions, we are led to place the eastern Himalayan boundary on the Nepal-Tibet watershed and the western boundary on the Zaskar range', this last range, the Zaskar, is made to fit into the chapter. It is all very complicated! There is the boundary between what the author calls the rivers of Tibet and Nepal, complicated by the fact that some of the Nepal rivers, such as the Arun, Trisuli Gandaki, and others, are Tibetan rivers also; this boundary is definitely in political Tibet. There is the boundary between Tibetan- and Aryan-speaking populations, which is somewhere, though extremely indefinite, in political Nepal. And there is the political boundary. The issue is still further complicated by a last- minute alteration in nomenclature in a footnote on page 105. Geographers will be wise if they use the term Tibet to mean political Tibet, in order to avoid such confusion of thought.

Chapter 12, on the Ladakh range and 'the Haramosh Ridge', is interesting and instructive, though not very clear. Geographers will be glad to note that the name 'Ladakh range' is no longer to be extended from the province of that name right across southern Tibet. The continuity of such a range has never been proved either geographically or structurally, and though, in the earlier stages of geographical investigation it may be permissible to postulate a range by linking up a line of high summits, the principle should be adopted with caution, when there are considerable gaps, as here. Sir Sidney gives a table of peaks observed to the south by Ryder and Wood on their journey up the Tsangpo in 1904 (pp. 104-5). Now when a triangulator follows the valley of a single long river, the points he fixes on either side naturally fall more or less into two alinements, though they do not necessarily lie on two primary ranges, as was suggested tentatively in 1907. A study of the modern maps shows that in many cases there is no pronounced trough between the southern alinement of triangulated points, hitherto called 'the Ladakh range,' and the Great Himalaya. Some of the points fixed by Wood are merely on the northern flanks of the Great Himalaya, others are not. Those shown on the map 62 K and the Great Himalayan summits are fifteen miles apart; in the Dhaulagiri region the two alinements are in contact; north of Mount Everest they are about sixty miles apart (map 71 L). The peaks fixed by Wood do not lie on the line shown as 'the Ladakh range' on the frontispiece to the first edition, which is renamed 'the Nepal-Tibet watershed' (though it is far from being the watershed)[1] in the second edition. On page 106 the author remarks on the striking parallelism. As drawn on the frontispiece the parallelism is certainly striking, but in fact neither the alinement of Wood's fixed points nor the watershed between the Tsangpo and the Ganges tributaries exhibits any parallelism to the Great Himalayan crest-zones. There is room for considerable research in the field on the existing maps and among existing literature of this subject. In the meanwhile it would be wise, not only to limit the use of the term 'Ladakh range' to the known section of the range in Ladakh, but also to treat any extension of it in either direction as quite hypothetical.

As with the Ladakh range, so with the Kailas: the old idea of a continuous 'Kailas range' has wisely been abandoned. Much of what was shown on the 1907 frontispiece by a continuous line is now shown more doubtfully by broken lines; the portion originally shown by a broken line has been omitted. It would perhaps have been still wiser to omit more. The old idea of a continuous Kailas range had apparently been suggested on the supposed alinement of triangulated points north of the Tsangpo, but Colonel Cowie was always sceptical about this so-called range, and proposed the term 'the south Tibet complex'. The name 'Kailas range' is now confined to the immediate neighbourhood of the sacred mountain. The trouble is that the long hypothetical 'Kailas range' and 'Ladakh range', which were most certainly only suggestions in 1907, were marked too confidently on the old charts, have been widely accepted, have been entered on our maps, and must unfortunately remain so until new editions are called for, and until sufficient publicity is given to the errors. It is easier to build up with facts, than to destroy an edifice of fallacy.

After these past experiences it will come as a shock to geographers to be asked to accept the suggestion of a continuous Karakoram range right across Tibet (Chapter 14) and to label it 'the Great Karakoram'. The principle of postulating a long, continuous, primary range, in the early stages of our knowledge, from an alinement of high triangulated peaks is here abandoned, and the new range suggested by Burrard has no more foundation than the existence of perhaps half a dozen passes in an area covering some twelve degrees of longitude, right across the least known part of Chang Tang.[2] There is still perhaps some difference of opinion as to the alinement of the main range of the Karakoram system in eastern Ladakh, where we now have good maps and considerable knowledge of the structure. Is it wise, considering the errors already made in connexion with the 'Ladakh range' and the 'Kailas range', to put forward so wild a guess on so little evidence? And having done so, to use this as an argument to support a hypothesis presented in 1907, before the modern survey of the Karakoram was made? The great Saser peaks, rising to over 24,000 and 25,000 feet, which have now been surveyed in detail, are known to lie on a great range continuous with the main range of the Karakoram. Both topographical and geological evidence point to this being in fact the continuation of the main range; yet the careful work of travellers and scientists in Baltistan and Ladakh is lightly set aside and this great range is stated to be a minor ridge, while a heavy black line is drawn on the frontispiece chart to indicate a primary range of the first magnitude across valleys, lakes, and plateaux, not yet scientifically examined, to link up with some doubtfully fixed passes, some hundreds of miles way. No scientific progress is possible by such methods.

Part III, which is entitled 'The Glaciers and Rivers of the Himalaya and Tibet', would have been more valuable if it had confined itself to these and kindred subjects. A fuller account of Himalayan climate by a qualified officer of the Indian Meteorological Department would have been a more useful introduction than the elementary and sketchy chapter on snowfall and rainfall with casual extracts from sportsmen's diaries (Chapter 17). There is no comprehensive account of the glacier systems of the Karakoram and Himalaya, and much of Chapter 18, which might have been occupied by this important branch of Himalayan geography, has been devoted to (a) a brief description of the survey of Kashmir and the Karakoram in 1855-65, mainly, it seems, in order to prove that by 1871 'the topographical survey of the Western Himalaya, conducted under Montgomerie, the accounts of its glaciers by Godwin Austen, and the examination of the southern portion by Medlicott have rendered our knowledge accurate and complete'; (b) a two-page criticism of Lord Conway and his works; (c) a short eulogy of the Workmans; (d) a long discussion of the names of the Saltoro pass and of Teram Kangri; (e) brief and not very accurate accounts of the surveys of recent travellers; (/) an appeal to Professor Dainelli to remember his Roman forefathers; (g) the views of certain military officers on the excellence of Survey of India maps; and (h) the briefest possible mention of the recent surveys in the Hindu Kush. The geography of the glaciers themselves and the many interesting problems of glaciology are barely mentioned. A summary of Himalayan climate, such as is suggested above, would also have afforded an opportunity for some mention of the natural vegetation. As it is, this important branch of geography has been ignored, though our knowledge of Himalayan forests is now very considerable.

Chapters XIX to XXV, dealing with the rivers of the Himalaya, are the best among the geographical sections of the book. The thread of geographical description is unfortunately destroyed by digressions into (a) the discovery of Mount Everest; (b) an unconvincing explanation of the symbols proposed for its 'satellites'; (c) a description of the first survey of Nepal; (d) the origin of the name ‘Kangchenjunga'- a renewal of the struggle that occupied some pages of Part I; (e) range bifurcations and conjunctions; and other matter irrelevant to the subjects of the chapters. There are also some inaccuracies: not by the greatest stretch of the imagination can it be said to-day that 'detailed knowledge of the glaciers of Kangchenjunga was gained from the maps made in 1874-84 ' !

Chapters 26 (The Central Asian Water-parting); 27 (The River- Gorges of the Himalaya); 28 (The Lakes of Tibet and Turkistan); 29 (On the Origin of Lakes) all contain matter of interest.

Part IV, 'The Geology of the Himalaya', follows closely on the lines of the admirable summary by Sir Henry Hayden in the first edition. If there is any comment to make, it is that there is too little mention of Dr. Heron's own valuable contributions, and those of his brother officers in the Geological Survey of India to-day. Such remarks as: 'Of late, however, physicists have begun to question the validity of Mr. Fisher's premises' (p. 283), while perfectly correct in 1907, are hardly true in 1934; and surely some of Mr. R. D. Oldham's work in the Simla-Chakrata area is inconsistent with modern interpretations. It is possible that Dr. Heron felt himself bound by the first three parts, or by the first edition, and so was unwilling to emphasize modern interpretations of structure; it may be that he does not agree with them. But a considered official summary of the observations described in recent Records and Memoirs of the Geological Survey would have been most valuable to European geographers and geologists, especially if it had appeared as the foundation for the geographical chapters. Probably such a summary would have required more leisure time than could be found by an officer still in harness; but here again it would have been worth while placing an officer on special duty to carry out the work. It is to be hoped that the 3rd edition of the Manual of the Geology of India, the last edition of which appeared as long ago as 1893, will review Himalayan geology and structure from the modern standpoint.

Controversy is generally most acute where facts are fewest.

[1] Chart XXIII of the new edition is identical with the same chart of the earlier edition and does not agree in many details with the new Charts XXVI, XXVII, XXVIII, XXX.

[2] The facts that the general 'grain' of the plateau is east and west, and that the few travellers who have crossed the plateau from north to south have crossed 'ranges' are no evidence at all of a continuity of a single great range.